USAMP AMD 408 –DIE FACE ENGINEERING FOR ADVANCED SHEET ... · USAMP AMD 408 – Die Face Engineering for Advanced Sheet Materials [email protected] February 28, 2008 This material

USAMP AMD 408 – Die Face Engineering for Advanced Sheet Materials

[email protected] February 28, 20081

Leader:Tom Stoughton

Administrator:Manish Mehta, TRC/NCMS

Presenter:Eric McCarty

February 28,2008

Prepared by T. B. Stoughton, General Motors on Feb 4, 2008

This presentation does not contain any proprietary or confidential information

USAMP AMD 408 – DIE FACE ENGINEERING FOR ADVANCED SHEET MATERIALS



Chrysler: Changqing DuFord Motor: Laurent Chappuis & Cedric Xia General Motors: Chung-Yeh Sa, Tom Stoughton,

Chuan-Tao Wang, & Siguang Xu

ALCOA: Edmund ChuUS Steel: Ming F. Shi & Ming Chen

Daimler: Tim LankeVolvo: Alf Andersson

TRC/NCMS Manish Mehta

Project Team




This material is based upon work supported by the Department of Energy National Energy Technology Laboratory under Award Number DE-FC05-02OR22910.This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.



Purpose of Work

Observed springback between stamped panel before and after trim.

Shape after trim

Shape after removal from stamping press

Springback Compensation:• If we know what shape we want the metal

product to be, what shape should the tool surface be?


Springback Prediction:• Sheet metal distorts due to elastic recovery

when forming stresses caused by tool contacts are removed.



Purpose of Work For Light-Weighting

State of the art of springback prediction and compensation technology relies too much on experience with mild steel.

Springback is more challenging for AHSS and aluminum alloys due to either higher forming stresses or lower elastic modulus.

These two factors impede the use of light-weighting metals.

Mild steel is the 11th Hour Solution during physical tryout to deal with difficult springback issues




Barriers and Approach To Springback PredictionChallenges

Complex Springback Modes: Side Wall Curl and Twist

Rigid Body Motion DoF and Panel Flexibility Complicate Measurement/Correlation Study

Solutions

Developed New Part to Exaggerate Springback Challenges

New Element Technology, Surface Contact/Continuity, and Material Models

Work with LSTC (LS-DYNA) to identify technology gaps and evaluate LSTC solutions

Improved Registration Methods and Fixture Simulation




Barriers and Approach To Springback CompensationChallenges

Maintaining Die Surface Continuity and Quality

Side Wall Curl and Twist compensation

Variability

Solutions

Define requirements and Conduct Benchmark Study (Think3D,ICEM,Tebis)

Combine Springback Minimization with Limited Compensation

Springback Minimization Based on Springback Prediction




Results: Improvements in Springback PredictionWorked with LSTC to develop list of developments in which improvement in FEM technology is considered an important factor in the accurate springback.

Assisted LSTC to implement solutions to these challenges.

Developed comprehensive experimental database for 8 light-weighting metals for calibration of advanced material models.

Developed Challenging Full-size Part and Common Die Tooling with Exaggerated Springback Challenge.

Evaluated LSTC improvements for accuracy in springback prediction.

Developed “Best Practice” of Existing and Improved Technology for Springback Prediction.


Comparison of predicted and measured springback along critical sections through the Common Die Part formed with the DP600 alloy.

Die Surface Scan

Old Technology Prediction

New Technology Prediction

Measured Springback of DP600



Validated Tool Compensation Via Springback PredictionDeveloped Standard for Surface Quality to be used in Springback Compensation

Completed Benchmark Study of Existing and Emerging Commercial Springback Compensation Software Solutions

Completed Pilot Test of Springback Prediction Based on Springback Prediction


Deviation From Design Intent of Panel Formed With Compensated Die



SummarySubstantial improvement in the reliability of springback prediction in draw forming has been achieved for DP600, DP780, TRIP780, and AA 5754-O.

Improvements made by LSTC will be rolled out in their manuals, training, and support program.

A “Best Practice” in the process of springback prediction has been developed and is implemented by the die engineering groups at GM, Ford, and Chrysler.

An extensive database for 8 aluminum and AHSS alloys has been created to calibrate advanced material models for cyclic loading process.

Commercial Springback Compensation Technology has been benchmarked and standards for surface quality have been developed.

Springback Compensation Based on Springback Prediction has been demonstrated for DP600 on a specially designed full sized die, modified from a production part to amplify springback issues.

Papers on the scientific contributions of the project are planned for the Numisheet 2008 Conference, and other forums and publications.

This project was completed in 4Q 2007 and the final report (of approximately 500 pages) will be available in 1Q 2008.




Plans for Next YearThere are still significant challenges facing the Virtual Manufacturing Process that are even more difficult for light-weigthing materials.

These challenges include:

• Simulation of Line Die Forming, which involve non-proportional loading histories that are outside the validated capability of even the most advanced material models available.

• Simulation of New Processes, such as high temperature forming, which is necessary for Boron steels, magnesium, and superplasticity of aluminum, add another dimension to material modeling, as well as an order-of-magnitude increase in material testing and database requirements.




Background Material and Details of Deliverables




Purpose of Work For Light-Weighting

Springback prediction/compensation is an immature technology that relies heavily on experience with the use of conventional mild steels.

More than 70% of physical tryout costs are due to springback compensation.

Due to the lower elastic modulus for aluminum and the higher forming stresses for HS steels, springback is significantly higher than it is for mild steels and more challenging.

Prior State of the Art




Development of the “Common Die” Tool to Exaggerate Springback Challenges

Y-Shape Rail Based on a Ford Production Part

Modified to contains features of interest to all OEM’s

Additional surface morphing involving twist to increase surface complexity

Original Ford Product Shape




Major Technical Obstacles to Accurate Springback Prediction

Cyclic loading has a significant impact on forming stresses

Conventional material models are not reliable for cyclic loading of anisotropic metals.

Sharp tool radii are responsible for generating most of the forming stresses

Conventional Contact Algorithms require too many elements on sharp tool radii leading to prohibitively costly simulations.

Normal (through-thickness) stresses are significant for high t/R conditions

Shell Elements do not consider non-zero normal stresses.Solid Elements are tool expensive for most production forming simulations




Improved Material Models

LEGEND

Grey – Die Face Surface Scan Brown – Formed and Trimmed Product Scan Data

Red – Predicted Springback Green – Predicted Springback

Using Old Material Model Using Modified Yoshida Model

Implemented/Validated/and Evaluated New Models for Improved Springback

Identified the F. Yoshida Model to be best in class for Aluminum

Modified the Yoshida Model to improve description of work hardening of AHSS alloys

Comparison of predicted and measured springback along critical sections through the Common Die Part formed with the DP600 alloy




Improved Material Models Require a New & Larger Database

Developed database, test matrix, and calibration procedures for advanced models for 8 aluminum and AHSS alloys. Test data include:

1) conventional uniaxial tension tests

2) compression, tension, and shear loading

3) single and multi-cycle loading with several strain increments

4) biaxial stress loading tests

Typical cyclic stress-strain response that requires advanced material models

Anisotropic evolution of yield surface shape with plastic strain

-400

-200

Sigma-2(Mpa)

0

200

400

600

800

0 200 400 600 800

Sig

ma-

1(M

pa)

0.00010.00020.00050.0010.0020.0050.010.020.030.040.050.06

DP600




Improved Surface Contact/ContinuityAreas of high curvature generate most of the

forming stresses and therefore play adominant role in springback

Conventional Surface Contact Algorithms require too many elements is areas of high curvature leading to high costs to obtain accurate springback results

A Smooth Contact Algorithm reduces number of elements without sacrificing accuracy

d

A

A’

B

Example:

Prediction of the springback angle for channel draw typically requires more than 4-10 elements on the die radius to obtain a converged solution.

The new Smooth Contact Algorithm is found to overcome this convergence problem with a course mesh (with only about a 20% increase in cost compared to conventional contact).

# of Elements 1 2 4 10

Springback Angle (deg) 19.9 21.2 20.8 21.5




Normal Stresses Are Significant when t/R is Large

The effect on deformation was noticeable, but was considered by the team to NOT be as significant as it was anticipated.

The new Shell Element is not adopted in the recommended “best practice”, with an exception perhaps in extreme cases for thick metals or very sharp radii.

A special Shell Element was developed by LSTC that computes the normal stress from the calculated normal surface contact pressure of associated nodes and factors its value in the constitutive relations.




Developed Requirements for Surface CompensationPrimary Manufacturing Challenges for Surface Compensation

Creation of undesirable die surface conditions:

eg. Backdrafts & oscillations

Magnification of surface gaps and slope discontinuities

beyond NC quality

Inordinate increase in surface data file size

Excessive manual repair to remedy surface issues

SolutionDeveloped Manufacturing Requirements for Die Compensation

Obtained Buy-In of Standards from Compensation Software Vendors

Evaluated Performance of Compensation Software Vendors




Evaluation of Surface Compensation StandardsBenchmark of Springback Compensation Technology Software

Participating vendors (Tebis, Think3D, and ICEM), others invited

Selected 6 automotive products providing complete design intent surfaces and the data for the expected springback in each case

No vendor satisfied all requirements of the standard

Tebis software was found to be leading the state of the art

Outstanding issue is the amount of manual repair required




Evaluation of Surface Compensation Technology

Tebis Compensation Solution Evaluated On Common Die Tool

Predict springback of DP600 using new prediction technology

Supply die compensation data based on springback result to Tebis

Repair (manually) the Tebis compensated die surface math data

Recut tool, assess tool surface quality, and stamp DP600 panels

Compare the surface of the trimmed part to the design intent

Benchmarking Compensation Technology -> ValidationThe benchmark served to establish the capability of commercial software to be able to morph surfaces to meet NC machine quality requirements

A final step is to evaluate the efficacy of the virtual springback prediction & compensation technologies as a system, on a production tool for an automotive application.




Evaluation of Surface Compensation Technology

Surface quality after manual repair was acceptable for production

The new product, formed from the compensated tool, was in most areas within 0.5 mm of design intent (shown in blue)

Design intent was not achieved in a few areas (where repairs were required to improve surface quality) --- it was later discovered that these manual modifications to the surface data were made in error due to the neglect of the springback compensation requirements

This outcome illustrates one of the reasons it is desired to minimize the amount of manual repair --- it’s not just the labor & time costs of the repair.

Area of Manual Repair

Deviation From Design Intent of Panel Formed With Compensated Die


Documents

USAMP AMD 408 –DIE FACE ENGINEERING FOR ADVANCED SHEET ... · USAMP AMD 408 – Die Face Engineering for Advanced Sheet Materials [email protected] February 28, 2008 This material